项目名称: 基于视觉伺服的AUV自主对接建模与非线性控制研究
项目编号: No.51279164
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 水利工程
项目作者: 高剑
作者单位: 西北工业大学
项目金额: 82万元
中文摘要: 本项目紧密结合我国建设海洋观测系统和水下无人作战系统对自主水下航行器(AUV)自主对接技术的迫切需求,在对AUV运动控制研究的基础上,提出AUV水下视觉伺服对接控制的技术概念,克服以往光学导引对接方法的不足,并提炼为多变量非线性控制问题开展理论和实验研究。借鉴机械手基于图像的视觉伺服控制技术,针对AUV三维空间运动、欠驱动约束、海洋环境扰动、对接目标运动等水下AUV对接的特点和研究难点深入研究,解决其中的可控性分析、非线性全局稳定对接控制律设计等关键问题,并分别利用移动机器人、无人水面船和水下航行器开展实验研究。
中文关键词: 水下航行器;视觉伺服;水下对接;模型预测控制;神经网络控制
英文摘要: Underwater docking control of autonomous underwater vehicles is one of the key techniques of the marine observation systems and the underwater unmanned troops. Borrowing from the idea of the visual servo control of robot manipulators, this project proposes the underwater visual servo docking control technology to overcome the disadvantages of traditional optical guidance methods, and carries out the theoretical and experimental researches as a multi-variable nonlinear control problem. The three dimensions motion of vehicle, the underactuated constraints, the underwater environment disturbances and the docking target's motion make this problem a challenging one. Based on the image-based visual servo control technology of manipulators, we will analyze the controllability of the system, design the nonlinear docking controller with global stability, and finally validate the control methods by the experiments with the wheeled robot, the autonomous surface vehicle and the underwater vehicle.
英文关键词: Underwater Vehicles;Visual Servoing;Underwater Docking;Model Predictive Control;Neural Network Control